[Libre-soc-bugs] [Bug 898] binutils svp64 objdump support
bugzilla-daemon at libre-soc.org
bugzilla-daemon at libre-soc.org
Tue Aug 30 17:35:46 BST 2022
https://bugs.libre-soc.org/show_bug.cgi?id=898
--- Comment #32 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
(In reply to Dmitry Selyutin from comment #31)
> These instructions do not appear in SVP64 CSV files
> ("^(?:LDST)?RM-(1P|2P)-.*?\.csv$"):
> ['b', 'bcctr', 'bctar', 'darn', 'dcbz', 'fabs', 'fabs.', 'fcpsgn',
> 'fcpsgn.', 'fmr', 'fmr.', 'fnabs', 'fnabs.', 'fneg', 'fneg.', 'hrfid',
> 'mcrxrx', 'mfcr', 'mfmsr', 'mfocrf', 'mtcrf', 'mtmsr', 'mtmsrd', 'mtocrf',
> 'rfid', 'sc', 'setvl', 'setvl.', 'svindex', 'svremap', 'svshape', 'td',
> 'tlbie', 'tw']
yes none of these are a surprise, the setvl svindex svremap svshape ones
are "management" (for which 64-bit variants haven't been defined).
mtmsr you absolutely cannot move more than one MSR (Machine Status Register)
so there is absolutely no point vectorising it.
rfid and sc are "return from interrupt" and "system call" respectively,
it is insanity to consider vectorising them.
branch (b) unconditionally is impossible to vectorise, how can you "branch"
to multiple locations?
others just haven't yet had an analysis done on them (fmr etc.)
--
You are receiving this mail because:
You are on the CC list for the bug.
More information about the libre-soc-bugs
mailing list